The major goals of the proposed research are to add to our knowledge of the capacities of young infants to perceive speech, to our understanding of the developing phonetic competence of young infants, and to a theory of speech perception by providing data bearing on the nature of the mechanisms that serve the perception of speech, that is, whether specialized for speech perception or general auditory processes, or both. More specifically, a series of experiments are planned with adult listeners that investigate the processes of integration (i.e., organization) of spatially separate sources of acoustic information, dichotically presented, that together form coherent linguistic units -- words in the case of adult listeners. These experiments are planned with the aim of describing the conditions under which integration occurs, that is, the conditions under which listeners treat information from different locations as the consequence of a single articulatory source. Additional experiments are planned that examine whether different linguistic units, segments as opposed to syllables, integrate more easily. In the same vein, experiments will test whether units that violate the presumed hierarchical structures of syllables affect processes of integration. Many of these experiments will be replicated with infants three and four months of age. The aim is to determine the breadth of the processes of integration and organization in the perception of speech by infants and whether these processes are influenced by the nature of the linguistic unit to be integrated. Evidence showing an effect of such units provides information about how speech is represented by infants and additional support for the contention that the acquisition of language, even in its earliest stages, is driven (in part) by processes of a linguistic nature. A second series of experiments with infants is proposed that will examine the structure of consonantal and vocalic categories in three- and four- month-old infants and in younger and older infants depending on the outcome of initial studies. Of particular concern is the age at which categorical structures can be demonstrated and whether these structures are determined by brief experience with various category exemplars. We are particularly interested in whether these structures are influenced by brief experience with nontypical exemplars. The existence of structures at an early age that are relatively insensitive to brief laboratory experience to various arrays of exemplars is compatible with the idea that both the categories of speech and their structures have a strong biological determination.